Increasingly, policy and programmatic responses to the HIV epidemic in the United States are challenged by complex patterns of comorbidity between major mental illness (MMI) and HIV disease. During the current grant period, we have developed and validated a range of methods that use New Jersey Medicaid data to address these issues, and have applied them to provide vitally important information on this relationship. In the continuation period, we will build on this initial work to develop and utilize recent data from five states (CA, FL, NJ, NY, and TX), from both Medicaid and Medicare. Analyses will provide a much more comprehensive understanding of HIV-MMI co-occurrence in time and space; on issues of quality and cost of care; on the complex interrelationships between treatment for psychiatric problems and HIV disease in this population; and on the role and impact of Medicaid and Medicare in this care at a time of many changes in these programs. Guided by a conceptual framework articulating hypothesized relationships among clinical and demographic characteristics, community resources, comorbidities and a range of outcomes, analyses will address evolving patterns of diagnosed comorbidity; patterns of treatment of MMI for the dually diagnosed as compared with those with HIV only and MMI only; and the impact of MMI and of treatment of comorbid mental disorders on HIV treatment and outcomes, including consistent use of antiretroviral medications, and use of hospitalization and other acute and long-term care services. We will analyze racial/ethnic and other socioeconomic disparities in care for persons with HIV, MMI or both, and identify intervening variables and targets for intervention. We will also examine variations in cost of care for beneficiaries living with HIV in relation to mental health comorbidity, with comparisons to cost of care for mentally ill individuals without HIV, and examine the role of Medicaid and Medicare for dually eligible individuals with these conditions. Results will provide knowledge that is useful at multiple levels. They will provide important information on the challenges of comorbid mental illness and HIV, and provide methodological tools for appropriate use of administrative data, and for modeling longitudinal services use and outcomes, that will be of general utility for mental health services research. [unreadable] [unreadable]